
**********************************************************************************************************************************
*How politicians downplay lower-educated citizens’ opinions
*Julie Sevenans & Stefaan Walgrave
*Accepted for publication in LSQ - February 2022
**********************************************************************************************************************************

use "LSQ_Replication data_How politicians downplay lower-educated citizens’ opinions.dta", clear

*Data preparation
alpha F2_1-F2_3
gen quality_evaluation = (F2_1 + F2_2 + F2_3) / 3

*MAIN PAPER
*Table  + predicted values in the text
reg quality_evaluation ib2.F1 
margins, at(F1=(1 2))
reg quality_evaluation ib2.F1 F2_6
margins, at(F2_6=(1(3)7))

*ONLINE APPENDIX

*Table A1.2 Conditions
tab F1

*Table A1.3 Balance test
//Three variables that are used for the balance test (gender, age, years of experience) 
//cannot be made public, due to promises that we made about confidentiality. 
//The variables would allow to personally identify MPs.
*tab F1, generate(group)
*logit group1 sex yob yofe
*logit group2 sex yob yofe
*logit group3 sex yob yofe
*logit group4 sex yob yofe

*Engagement with the stimulus/questions
sum TF2_PS

*Table A2.1 Descriptives
sum quality_evaluation F2_1-F2_3 F2_6

*Figure A3.1 Opinions by party
tab V4, sum(F2_6)
graph hbar F2_6 if V4!=7, over(V4, sort(1) descending) ///
	exclude0 yscale(range(1 7)) ylabel(1(1)7) ///

*Table A3.1 Explaining own opinion
reg F2_6 ib2.F1 ib5.V4

*Table A4.1 Analysis with party dummies
reg quality_evaluation ib2.F1 ib5.V4
reg quality_evaluation ib2.F1 F2_6 ib5.V4



